Augmented Ebola Optimization Algorithm for Anomaly-Based Intrusion Detection System

Abstract

The exponential growth of information technology results in a higher proliferation of internet-connected products. In addition, there was a rise in the Number of network attacks. An intrusion detection system (IDS) is considered one of the most efficient security solutions. This research presents a novel anomaly-based intrusion detection system (IDS) model that utilizes the Ebola Optimization Algorithm (EOA) technique for increased performance. The EOA algorithm is hybridized with the Prairie Dog Algorithm (PD) to enhance the performance of the EOA algorithm, called EOA-PD. The EOA-PD technique is used as a feature selection mechanism to determine the most pertinent elements from the dataset that lead to a high level of classification accuracy. In addition, a support vector machine is utilized to assess the effectiveness of the chosen features in correctly forecasting the attacks. Furthermore, The NSL-KDD dataset is used to showcase the efficacy of the suggested methodology across several attack scenarios.